Spatio-temporal heterogeneity and driving mechanism of ecosystem water use efficiency in the Loess Plateau, China

The Loess Plateau, China The trade-off between carbon uptake and water loss, characterized by ecosystem water use efficiency (WUE), deeply influences ecosystem sustainability. In this study, WUE was estimated based on GPP from the Moderate Resolution Imaging Spectroradiometer (MODIS) product and ET simulated by the Priestley Taylor Jet Propulsion Laboratory (PT-JPL) model. The emerging hot spot analysis (EHSA) was used to comprehensively examine the spatio-temporal heterogeneity of WUE. Additionally, the Geodetector model was employed to identify the main drivers of WUE and quantify the interactive effects of drivers on WUE, focusing on different vegetation types. Results indicated obvious spatial heterogeneity of WUE under different vegetation and hydro-climatic conditions. Significant increases in WUE occurred in sub-humid cropland and grassland areas that have experienced large-scale ecological restoration and agricultural intensification. By contrast, slight decreases in WUE were observed in semi-arid grassland areas, some of which were accompanied by intensifying cold spots. It was noteworthy that WUE in some hot spots with excessive vegetation coverage also showed a slight downward trend. Further, the WUE pattern in cropland/forests/grassland was dominated by the interaction of vegetation coverage (characterized by the Normalized Difference Vegetation Index, NDVI) and precipitation/air temperature/vapor pressure deficit, with NDVI playing a leading role and hydro-climatic drivers playing a supporting role. • Spatio-temporal heterogeneity of WUE is examined considering spatial clustering. • NDVI has the strongest explanatory power for the spatial heterogeneity of WUE. • Interactive effects between drivers show bivariate and nonlinear enhancement.